Subversion Repositories openrisc

//////////////////////////////////////////////////////////////////////

////                                                              ////

////  or1200_fpu_post_norm_addsub                                 ////

////                                                              ////

////  This file is part of the OpenRISC 1200 project              ////

////  http://opencores.org/project,or1k                           ////

////                                                              ////

////  Description                                                 ////

////  post-normalization entity for the addition/subtraction unit ////

////                                                              ////

////  To Do:                                                      ////

////                                                              ////

////                                                              ////

////  Author(s):                                                  ////

////      - Original design (FPU100) -                            ////

////        Jidan Al-eryani, jidan@gmx.net                        ////

////      - Conv. to Verilog and inclusion in OR1200 -            ////

////        Julius Baxter, julius@opencores.org                   ////

////                                                              ////

//////////////////////////////////////////////////////////////////////

//

//  Copyright (C) 2006, 2010

//

//      This source file may be used and distributed without        

//      restriction provided that this copyright statement is not   

//      removed from the file and that any derivative work contains 

//      the original copyright notice and the associated disclaimer.

//                                                           

//              THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY     

//      EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED   

//      TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS   

//      FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL THE AUTHOR      

//      OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,         

//      INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES    

//      (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE   

//      GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR        

//      BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF  

//      LIABILITY, WHETHER IN  CONTRACT, STRICT LIABILITY, OR TORT  

//      (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT  

//      OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE         

//      POSSIBILITY OF SUCH DAMAGE. 

//

module or1200_fpu_post_norm_addsub

  (

   clk_i,

   opa_i,

   opb_i,

   fract_28_i,

   exp_i,

   sign_i,

   fpu_op_i,

   rmode_i,

   output_o,

   ine_o

   );

   parameter FP_WIDTH = 32;

   parameter MUL_SERIAL = 0; // 0 for parallel multiplier, 1 for serial

   parameter MUL_COUNT = 11; //11 for parallel multiplier, 34 for serial

   parameter FRAC_WIDTH = 23;

   parameter EXP_WIDTH = 8;

   parameter ZERO_VECTOR = 31'd0;

   parameter INF = 31'b1111111100000000000000000000000;

   parameter QNAN = 31'b1111111110000000000000000000000;

   parameter SNAN = 31'b1111111100000000000000000000001;

   input     clk_i;

   input [FP_WIDTH-1:0] opa_i;

   input [FP_WIDTH-1:0] opb_i;

   input [FRAC_WIDTH+4:0] fract_28_i;

   input [EXP_WIDTH-1:0]  exp_i;

   input                  sign_i;

   input                  fpu_op_i;

   input [1:0]             rmode_i;

   output reg [FP_WIDTH-1:0]       output_o;

   output reg             ine_o;

   wire [FP_WIDTH-1:0]     s_opa_i;

   wire [FP_WIDTH-1:0]     s_opb_i;

   wire [FRAC_WIDTH+4:0] s_fract_28_i;

   wire [EXP_WIDTH-1:0] s_exp_i;

   wire s_sign_i;

   wire s_fpu_op_i;

   wire [1:0] s_rmode_i;

   wire [FP_WIDTH-1:0] s_output_o;

   wire s_ine_o;

   wire s_overflow;

   wire [5:0] s_zeros;

   reg [5:0] s_shr1;

   reg [5:0] s_shl1;

   wire s_shr2, s_carry;

   wire [9:0] s_exp10;

   reg [EXP_WIDTH:0] s_expo9_1;

   wire [EXP_WIDTH:0] s_expo9_2;

   wire [EXP_WIDTH:0] s_expo9_3;

   reg [FRAC_WIDTH+4:0] s_fracto28_1;

   wire [FRAC_WIDTH+4:0] s_fracto28_2;

   wire [FRAC_WIDTH+4:0] s_fracto28_rnd;

   wire s_roundup;

   wire s_sticky;

   wire s_zero_fract;

   wire s_lost;

   wire s_infa, s_infb;

   wire s_nan_in, s_nan_op, s_nan_a, s_nan_b, s_nan_sign;

   assign s_opa_i = opa_i;

   assign s_opb_i = opb_i;

   assign s_fract_28_i = fract_28_i;

   assign s_exp_i = exp_i;

   assign s_sign_i = sign_i;

   assign s_fpu_op_i = fpu_op_i;

   assign s_rmode_i = rmode_i;

   // Output Register

   always @(posedge clk_i)

     begin

        output_o <= s_output_o;

        ine_o <= s_ine_o;

     end

   //*** Stage 1 ****

   // figure out the output exponent and how much the fraction has to be 

   // shiftd right/left

   assign s_carry = s_fract_28_i[27];

   reg [5:0] lzeroes;

   always @(s_fract_28_i)

     casez(s_fract_28_i[26:0])   // synopsys full_case parallel_case

       27'b1??????????????????????????: lzeroes = 0;

       27'b01?????????????????????????: lzeroes = 1;

       27'b001????????????????????????: lzeroes = 2;

       27'b0001???????????????????????: lzeroes = 3;

       27'b00001??????????????????????: lzeroes = 4;

       27'b000001?????????????????????: lzeroes = 5;

       27'b0000001????????????????????: lzeroes = 6;

       27'b00000001???????????????????: lzeroes = 7;

       27'b000000001??????????????????: lzeroes = 8;

       27'b0000000001?????????????????: lzeroes = 9;

       27'b00000000001????????????????: lzeroes = 10;

       27'b000000000001???????????????: lzeroes = 11;

       27'b0000000000001??????????????: lzeroes = 12;

       27'b00000000000001?????????????: lzeroes = 13;

       27'b000000000000001????????????: lzeroes = 14;

       27'b0000000000000001???????????: lzeroes = 15;

       27'b00000000000000001??????????: lzeroes = 16;

       27'b000000000000000001?????????: lzeroes = 17;

       27'b0000000000000000001????????: lzeroes = 18;

       27'b00000000000000000001???????: lzeroes = 19;

       27'b000000000000000000001??????: lzeroes = 20;

       27'b0000000000000000000001?????: lzeroes = 21;

       27'b00000000000000000000001????: lzeroes = 22;

       27'b000000000000000000000001???: lzeroes = 23;

       27'b0000000000000000000000001??: lzeroes = 24;

       27'b00000000000000000000000001?: lzeroes = 25;

       27'b000000000000000000000000001: lzeroes = 26;

       27'b000000000000000000000000000: lzeroes = 27;

     endcase

   assign s_zeros = s_fract_28_i[27] ? 0 : lzeroes;

   // negative flag & large flag & exp          

   assign s_exp10 = {2'd0,s_exp_i} + {9'd0,s_carry} - {4'd0,s_zeros};

   always @(posedge clk_i)

     begin

        if (s_exp10[9] | !(|s_exp10))

          begin

             s_shr1 <= 0;

             s_expo9_1 <= 9'd1;

             if (|s_exp_i)

               s_shl1 <= s_exp_i[5:0] - 6'd1;

             else

               s_shl1 <= 0;

          end

        else if (s_exp10[8])

          begin

             s_shr1 <= 0;

             s_shl1 <= 0;

             s_expo9_1 <= 9'b011111111;

          end

        else

          begin

             s_shr1 <= {5'd0,s_carry};

             s_shl1 <= s_zeros;

             s_expo9_1 <= s_exp10[8:0];

          end // else: !if(s_exp10[8])

     end // always @ (posedge clk_i)

   //-

   // *** Stage 2 ***

   // Shifting the fraction and rounding

   always @(posedge clk_i)

     if (|s_shr1)

       s_fracto28_1 <= s_fract_28_i >> s_shr1;

     else

       s_fracto28_1 <= s_fract_28_i << s_shl1;

   assign s_expo9_2 = (s_fracto28_1[27:26]==2'b00) ?

                      s_expo9_1 - 9'd1 : s_expo9_1;

   // round

   //check last bit, before and after right-shift

   assign s_sticky = s_fracto28_1[0] | (s_fract_28_i[0] & s_fract_28_i[27]);

   assign s_roundup = s_rmode_i==2'b00 ?

                      // round to nearset even

                      s_fracto28_1[2] & ((s_fracto28_1[1] | s_sticky) |

                                         s_fracto28_1[3]) :

                      s_rmode_i==2'b10 ?

                      // round up

                      (s_fracto28_1[2] | s_fracto28_1[1] | s_sticky) & !s_sign_i:

                      s_rmode_i==2'b11 ?

                      // round down

                      (s_fracto28_1[2] | s_fracto28_1[1] | s_sticky) & s_sign_i :

                      // round to zero(truncate = no rounding)

                      0;

   assign s_fracto28_rnd = s_roundup ?

                           s_fracto28_1+28'b0000_0000_0000_0000_0000_0000_1000 :

                           s_fracto28_1;

   // ***Stage 3***

   // right-shift after rounding (if necessary)

   assign s_shr2 = s_fracto28_rnd[27];

   assign s_expo9_3 = (s_shr2 &  s_expo9_2!=9'b011111111) ?

                      s_expo9_2 + 9'b000000001 : s_expo9_2;

   assign s_fracto28_2 = s_shr2 ? {1'b0,s_fracto28_rnd[27:1]} : s_fracto28_rnd;

   ////-

   assign s_infa = &s_opa_i[30:23];

   assign s_infb = &s_opb_i[30:23];

   assign s_nan_a = s_infa &  (|s_opa_i[22:0]);

   assign s_nan_b = s_infb &  (|s_opb_i[22:0]);

   assign s_nan_in = s_nan_a | s_nan_b;

   // inf-inf=Nan

   assign s_nan_op = (s_infa & s_infb) &

                     (s_opa_i[31] ^ (s_fpu_op_i ^ s_opb_i[31]));

   assign s_nan_sign = (s_nan_a & s_nan_b) ? s_sign_i :

                       s_nan_a ?

                       s_opa_i[31] : s_opb_i[31];

   // check if result is inexact;

   assign s_lost = (s_shr1[0] & s_fract_28_i[0]) |

                   (s_shr2 & s_fracto28_rnd[0]) | (|s_fracto28_2[2:0]);

   assign s_ine_o = (s_lost | s_overflow) & !(s_infa | s_infb);

   assign s_overflow = s_expo9_3==9'b011111111 & !(s_infa | s_infb);

   // '1' if fraction result is zero

   assign s_zero_fract = s_zeros==27 & !s_fract_28_i[27];

   // Generate result

   assign s_output_o = (s_nan_in | s_nan_op) ?

                       {s_nan_sign,QNAN} :

                       (s_infa | s_infb) | s_overflow ?

                       {s_sign_i,INF} :

                       s_zero_fract ?

                       {s_sign_i,ZERO_VECTOR} :

                       {s_sign_i,s_expo9_3[7:0],s_fracto28_2[25:3]};

endmodule // or1200_fpu_post_norm_addsub